Further Results on Robust Variance-Constrained Filtering for Uncertain Stochastic Systems with Missing Measurements
Further Results on Robust Variance-Constrained Filtering for Uncertain Stochastic Systems with Missing Measurements
复制标题
DOI：
10.1007/s00034-010-9178-4
复制
发表时间：
2010-03
期刊：
影响因子：
--
通讯作者：
Yijing Wang;Z. Zuo
中科院分区：
文献类型：
--
作者：
Yijing Wang;Z. Zuo
本文重新研究了具有观测缺失的不确定离散随机系统的鲁棒滤波问题。系统的测量值在任何采样时间都可能不可用。我们的目标是设计一个新的滤波器，使得滤波过程的误差状态是均方有界的。此外，每个状态的估计误差的稳态方差不超过个人规定的上限，所有允许的不确定性和所有可能的不完全观测。它表明，鲁棒滤波器的设计可以进行直接求解一组线性矩阵不等式。在推导过程中，不需要对系统矩阵A作非奇异性假设，也不需要用不等式来处理不确定性。因此，预期可以获得较不保守的条件。通过一个示例说明了新方法的优点。
This paper revisits the problem of robust filtering for uncertain discrete-time stochastic systems with missing measurements. The measurements of the system may be unavailable at any sample time. Our aim is to design a new filter such that the error state of the filtering process is mean-square bounded. Furthermore, the steady-state variance of the estimation error of each state does not exceed the individual prescribed upper bound subject to all admissible uncertainties and all possible incomplete observations. It is shown that the design of a robust filter can be carried out by directly solving a set of linear matrix inequalities. The nonsingular assumption on the system matrixAand the inequality which is used to handle the uncertainties are not necessary in the derivation process of our results. Thus, it is expected that a less conservative condition can be obtained. The advantage of the new method is demonstrated via an illustrative example.